nstx-u 5 year plan for pedestal, scrape-off layer and divertor physics

NSTX-U 5 Year Plan for Pedestal, Scrape-off Layer and Divertor Physics

V. A. Soukhanovskii (LLNL)A. Diallo, R. Maingi, C. S. Chang (PPPL)

for the NSTX-U Research Team

NSTX-U PAC-33 MeetingPPPL – B318

February 19-21, 2013

NSTX-U Supported by

Culham Sci CtrYork U

Chubu UFukui U

Hiroshima UHyogo UKyoto U

Kyushu UKyushu Tokai U

NIFSNiigata UU Tokyo

JAEAInst for Nucl Res, Kiev

Ioffe InstTRINITI

Chonbuk Natl UNFRI

KAISTPOSTECH

Seoul Natl UASIPP

CIEMATFOM Inst DIFFER

ENEA, FrascatiCEA, Cadarache

IPP, JülichIPP, Garching

ASCR, Czech Rep

Coll of Wm & MaryColumbia UCompXGeneral AtomicsFIUINLJohns Hopkins ULANLLLNLLodestarMITLehigh UNova PhotonicsOld DominionORNLPPPLPrinceton UPurdue USNLThink Tank, Inc.UC DavisUC IrvineUCLAUCSDU ColoradoU IllinoisU MarylandU RochesterU TennesseeU TulsaU WashingtonU WisconsinX Science LLC

NSTX-U NSTX-U PAC-33 – V. A. Soukhanovskii, NSTX-U 5 Year Plan for Pedestal, SOL, and Divertor Physics 2 of 18

Boundary Physics program in NSTX-U contributes to critical research areas for ITER/tokamaks and STs

High-level goals for NSTX-U 5 year plan • Demonstrate stationary non-inductive operation• Access reduced n* and high-b • Develop and understand non-inductive start-up/ramp-up• Develop and utilize high-flux-expansion snowflake

divertor and radiative detachment for mitigating very high heat fluxes

• Begin to assess high-Z PFCs + liquid lithium



• Boundary Physics Thrusts (and outline of the talk)– BP-1: Assess and control pedestal structure, edge transport and

stability• Pedestal structure, transport and turbulence studies• ELM characterization and control

– BP-2: Assess and control divertor heat and particle fluxes• SOL transport and turbulence, impurity transport• Divertor heat flux mitigation with impurity seeding and divertor geometry

• This talk: Recent NSTX Boundary Physics progress, Goals and Plans for Pedestal, ELM, SOL and divertor research in 2014-2018


Planned NSTX-U facility upgrades enable access to new parameter space and unique capabilities

Planned upgrades(NSTX NSTX-U )

Operationsyear

available

Boundary Physics area

PNBI = 5 12 MW (5 s) 7.5 15 MW (1.5 s)

1-2 Pedestal structure and ELM stability, L-H, divertor heat flux (12 15-20 MW/m2)P/R ~ 10 20P/S ~ 0.2 0.4

Ip = 1.3 2 MABt = 0.5 1 T

1-2 L-H transition, pedestal structure and stability, SOL width, divertor heat flux

Pulse length 1.5 5-10 s 1-2 Steady-state divertor heat flux mitigation, density and impurity control

Axisymmetric PF (divertor) coils PF1A, 1B, 1C, 2L

1-3 Plasma shaping, L-H, divertor configuration control

Non-axisymmetric control coils

3 ELM control and pedestal transport

Divertor cryogenic panel 3 Pedestal stability, density control, radiative divertor with impurity seeding

Molybdenum plasma-facing components

2 Core and pedestal impurity density, divertor heat transport regimes





– BP-2: Assess and control divertor heat and particle fluxes• SOL transport and turbulence, impurity transport• Divertor heat flux mitigation with impurity seeding and divertor geometry


Pedestal studies focus on testing applicability of peeling ballooning and kinetic ballooning to limit pedestal heights widths, and gradients

• Peeling ballooning limits on pedestal height consistent with ELMy/ELM-free operation- Challenge: applicability of

diamagnetic stabilization- Lithium and 3-D fields used to

manipulate profiles• Kinetic ballooning being tested as

a mechanism to limit width- Scaling with bpol

ped stronger in STs than higher R/a

- No evidence yet of KBM fluctuation

BP-1

Diagnostics: 30 point MPTS (NSTX) 42 point MPTS (NSTX-U)

C-Mod

0.1

0.04

0.08

0.12

0.16

0.00

Ped

esta

l wid

th (y

N)

0.3 0.5 0.7

~ 0.08 (bqped)0.5

0.4 (bqped)1.05 NSTX

DIII-D


Understand role of different microinstabilities for different transport channels for pedestal control

• TEM / KBM modes appear dominant in near-edge region of NSTX pedestal– GENE: ITG/KBM-TEM hybrid modes and microtearing– GS2: hybrid TEM/KBM dominant

• No lithium: microtearing modes linearly unstable• Lithium: microtearing stable, TEM unstable, ETG unstable

near separatrix

• Measured correlation lengths at pedestal top are consistent with theory (XGC1)– Spatial structure exhibits ion-

scale microturbulence – Compatible with ITG modes

and/or KBM

Diagnostics (NSTX-U): 42 point MPTS, reflectometry, 48 point BES

BP-1

3 cm 2 – 4 cm(reflecto-metery)

11 cm 10 – 14 cm(BES)

Theory(non-linear XGC1 code)

R = 1.38m

Experiment

80% - 99% ELM cycle

radial

poloidal

139047

LithiumNo lithium


ELM evolution and divertor/wall heat flux studies in NSTX-U to focus on acceptable ELM heat flux scenarios

• Many ELM regimes observed in NSTX: Type I, III, V, mixed, (and also Type II in narrow operational windows)- Phenomenology dependent on n*ped, PSOL,

Ip, shaping

• ELM control techniques - Lithium granule injector (tested at EAST)- Lithium evaporation- Enhanced Pedestal (ELM-free) H-mode- 3-D fields (n=3 RMP)

• Heat flux from Type I & III ELMs measured to be very high; small Type V ELMs have low heat flux- Type I ELMs triggered with 3-D fields can

reduce per ELM energy loss and peak heat flux, but slower than 1/nELM

- Toroidal asymmetry increases with ELM peak heat flux

Type I

Type III

Type V

Mixed


Plans for pedestal transport, turbulence and ELM control research

• Year 1 of 5 Year Plan– Continue cross-machine comparison of pedestal structure with DIII-D and Alcator C-Mod– Continue gyro-kinetic modeling of electromagnetic turbulence

• Years 2-3 of 5 Year Plan– Pedestal structure and turbulence vs engineering and physics parameters – ELM control with lithium coatings, lithium granule injector, and 3D fields\– Access to new operational regimes (enhanced pedestal H-mode, I-mode)– Initiate assessment of SOL current generation and impact on ELM models and control– Re-establish parameter regime for observation of edge harmonic oscillations

• Years 4-5 of 5 Year Plan – ELM control with off-midplane non-axisymmetric control coils– Compare pedestal structure and turbulence with edge transport models, XGC0 and XGC1– Assess cryopump and molybdenum PFC impact on pedestal collisionality and control

Diagnostics: MPTS, CHERS, BES, reflectometry, GPI





– BP-2: Assess and control divertor heat and particle fluxes• SOL transport and turbulence, impurity transport• Divertor heat flux mitigation with impurity seeding and divertor

geometry


NSTX-U data will help reduce SOL width scaling uncertainties for ST-FNSF and ITER

• SOL width scaling– In NSTX: lq

mid ~ Ip-1.6, independent of PSOL and Bt

• For NSTX-U (2 MA): lqmid = 3±0.5 mm

– Multi-machine database (Eich IAEA FEC 2012):lq

mid (mm) = (0.63+/-0.08)x Bpol,MP-1.19

• For NSTX-U (Bp~0.55 T): lqmid ~ 1.3 mm

• Divertor peak heat flux scaling– qpeak~ PSOL and qpeak ~ Ip – For NSTX-U qpeak ~ 20-30 MW/m2

Diagnostics: IR cameras, MPTS, Langmuir probes– Divertor Thomson scattering (incremental) valuable for

pressure/temperature balance between midplane, x-point, and divertor regions for heat transport studies

BP-2


SOL turbulence studies will help assess relative importance of turbulent and drift-based transportBP-2

GPI SOLT

• Comparison of lSOL with SOL models– Parallel transport: conductive/convective, cross-field :

collisional / turbulent / drift– Goldston drift-based model

• lSOL ~ (2a/R) rq, for NSTX-U: lqmid ~ 6 mm

– Exploring mechanisms setting steep pressure gradient region and connection to SOL width

– XGC0 (drift-kinetic): lqmid ~ Ip

-1.0

– SOLT (fluid turbulence): Ip scaling is weaker than observed

• Comparison of GPI, BES data and simulations elucidate on SOL transport and L-H transition– Blob formation, motion, interaction with sheared flows– Role of magnetic X-point geometry– Effects of 3-D fields on turbulence– Role of atomic physics

Diagnostics: MPTS, GPI – multiple views, BES, reflectometry, Langmuir probes


Snowflake divertor geometry has benefits over standard divertor

Standard divertor Snowflake-minus Exact snowflake • Second-order null

– Bp ~ 0 and grad Bp ~ 0• Significant divertor heat

flux reduction due to geometry confirmed in NSTX and DIII-D

BP-2

• Outstanding questions– Magnetic control of up-down symmetric snowflake– Effect on pedestal and ELMs– ELM convective heat transport in null point region– SOL and divertor turbulence disconnection– X-point transport and drifts– Compatibility with cryopumping


Multi-machine snowflake divertor studies validate the snowflake divertor concept for NSTX-U and ST-FNSF

Snowflake

BP-2

Snowflake

• Snowflake divertor in NSTX (2009-2010)– High confinement, core impurity reduction– Destabilized ELMs suppressed by lithium– Reduced heat flux by x 2-5 between and during ELMs– Outer strike point radiatively detached

• Initial DIII-D experiments (2012)– Used NSTX-style snowflake control scenario– Long pulse (3s) snowflake, high tE, H(89P) ≥ 2 – Heat flux reduced x 2-3, strike point attached– Reduced ΔWELM and qELM

pk (with D2 seeding)


Simulations of snowflake divertor configuration for NSTX-U yield optimistic projections

• Significant geometry effects• Multi-fluid edge transport

model (UEDGE) – Heat flux reduction with 4% carbon

by geometry and radiation in outer strike point• for 12 MW NBI case from 15 MW/m2 to

2-3 MW/m2

– Inner divertor detached – Particle removal by cryopump

results in reduced radiation • But still significant heat flux reduction

due to geometry

Standard Snowflake

LX (m) 5 17

fexp 20 60-100

Vdiv (m3) 0.070 0.128

BP-2


Impurity-seeded radiative divertor with feedback control is planned for NSTX-UBP-2

• In NSTX, heat flux reduction in radiative divertor compatible with H-mode confinement was demonstrated with D2 or CD4 puffing

• Seeded impurity choice dictated by Zimp and PFC– Li/C PFCs compatible with D2, CD4, Ne, Ar seeding– UEDGE simulations show Ar most effective

• Too much argon causes radiation collapse of pedestal • Feedback control of divertor radiation via impurity particle

balance control– Cryopump for particle removal + divertor gas

injectors– Real-time control signal diagnostics identified for

NSTX-U• PFC temperature via IR thermography or thermocouple• Thermoelectric current between inner and outer divertor• Impurity VUV spectroscopy or bolometry• Neutral gas pressure or electron-ion recombination rate


SOL and divertor studies will focus on model validation and heat flux mitigation

• Year 1– Continue analysis of SOL width database and comparison with models– Collaboration with DIII-D on snowflake and radiative divertor experiments

• Years 2-3 – Establish SOL width and divertor database vs. engineering and physics parameters – Re-establish edge turbulence measurements (GPI, BES, cameras, probes)– Initial radiative divertor experiments with D2, CD4 and Ar seeding and lithium– Develop snowflake divertor magnetic control and assess pedestal stability, divertor

power balance, turbulence, 3D fields as functions of engineering parameters– Comparison with multi-fluid and gyro-kinetic models

• Years 4-5 – 3D edge/SOL turbulence structure vs. edge plasma parameters, magnetic divertor

geometry, and 3D fields and comparison with 3-D turbulence models– Snowflake divertor with cryopumping and molybdenum PFCs in H-mode scenarios– Radiative divertor with feedback control of impurity seeding and cryo-pumping

BP-2

Diagnostics: MPTS, CHERS, BES, GPI, IR cameras, Langmuir probes, spectroscopy• Divertor Thomson scattering (incremental) highly beneficial for model validation


Boundary program focuses on advancing pedestal physics, power and particle handling with new NSTX-U capabilities

• Pedestal physics: confirm consistency with peeling ballooning, and test applicability of kinetic ballooning - Use lithium conditioning and 3-D fields as a way to manipulate the

density and pressure profile• Power and particle handling: further develop snowflake and

radiative divertors- Test key predictions of snowflake configuration, and evaluate

synergy with radiative divertors, graphite and molybdenum plasma facing components

- Evaluate compatibility with cryopumping• Planned research aims at providing pedestal and divertor

physics basis for ST-FNSF


Backup


Thrust BP-1: Assess, optimize, and control pedestal structure, edge transport and stability

• Characterize pedestal structure using the extended parameter range of NSTX-U – Measure turbulence and zonal flow dynamics with beam emission

spectroscopy, reflectometry, and gas-puff imaging diagnostics– Assess maximum achievable pedestal height, variation in pedestal

structure with increased field, current, power and shaping

• Increase control of pedestal transport and stability– Density profile modification with improved impurity and density control– ELM triggering/suppression with 3D fields from mid-plane and off-

midplane partial-NCC coil-sets and triggering with Li granule injection– Optimize enhanced-pedestal H-modes with above tools– Excite edge harmonic oscillations to increase particle x-port (increm.)


Thrust BP-2: Assess and control divertor heat and particle fluxes

• Investigate SOL heat and particle transport and turbulence– Extend heat flux width studies to lower n, higher BT, IP, and PSOL

– Compare data to multi-fluid turbulence and gyro-kinetic models • Investigate the snowflake divertor for power and particle

control– Develop magnetic control of conventional and snowflake divertors– Study steady-state & transient heat & particle transport, divertor loads

• Vary magnetic balance, feedback-controlled impurity seeding rate• Develop highly-radiating boundary solutions with feedback

control, determine divertor detachment operating window• Validate cryo-pump physics assumptions, perform initial

density control studies, determine range of accessible density• Assess impact of high-Z metal divertor PFCs on H-mode

confinement, impurity accumulation, power exhaust


Divertor Thomson Scattering system would significantly enhance NSTX-U Boundary research capabilities

• Physics contributions– ELM and inter-ELM divertor transport– SOL width scaling, role of X-point heat transport– Radiative detachment model validation– Snowflake divertor properties, incl. X-pt bp measurements– Divertor plasma-surface interaction and impurity transport

studies, erosion rates via spectroscopy • A conceptual geometry identified for NSTX-U• Conceptual system design underway


Edge Harmonic Oscillations

EHO’s limit density rise in QH modes on DIII-D. They are observed on NSTX with Li, but don’t seem to limit density.

Rob Goldston, Eric Fredrickson, Neal Crocker, Mike Jaworski

Mirnov

Langmuir Probes

However EHOs do affect the edge and SOL strongly.

Reflectometry


Driving Edge Harmonic OscillationsKey to Active Edge Control?

Jong-Kyu Park

MHD calculations indicate we can amplify edge kinks by driving HHFW antenna straps at audio frequencies.

Can this give us external control over edge pressure gradient(and so ELMs) and/or the SOL width?


Snowflake divertor geometry has benefits over standard X-point divertor geometry

• Predicted geometry properties in snowflake divertor (cf. standard divertor)– Increased edge shear :ped. stability– Add’l null: H-mode power threshold, ion loss– Larger plasma wetted-area Awet : reduce qdiv

– Four strike points : share qII

– Larger X-point connection length Lx : reduce qII

– Larger effective divertor volume Vdiv : incr. Prad , PCX

D. D. Ryutov, PoP 14, 064502 2007


Snowflake divertor geometry takes advantage of second-order poloidal field null properties

Standard divertor Snowflake-minus Exact snowflake

• Predicted geometry properties in snowflake divertor (cf. standard divertor)– Increased edge shear :ped. stability– Add’l null: H-mode power threshold, ion loss– Larger plasma wetted-area Awet : reduce qdiv

– Four strike points : share qII

– Larger X-point connection length Lx : reduce qII

– Larger effective divertor volume Vdiv : incr. Prad , PCX

• Second-order null– Bp ~ 0 and grad Bp ~ 0 (Cf. first-

order null: Bp ~ 0)• Obtained with existing

divertor coils (min. 2)• Exact snowflake

topologically unstable

BP-2


NSTX: good H-mode confinement properties and core impurity reduction obtained with snowflake divertor

0.8 MA, 4 MW H-mode k=2.1, d=0.8 Core Te ~ 0.8-1 keV, Ti ~ 1 keV bN ~ 4-5 Plasma stored energy ~ 250 kJ H98(y,2) ~ 1 (from TRANSP) ELMs

Suppressed in standard divertor H-mode via lithium conditioning

Re-appeared in snowflake H-mode

Core carbon reduction due to• Type I ELMs• Edge source reduction

• Divertor sputtering rates reduced due to partial detachment


Impulsive heat loads due to Type I ELMs are mitigated in snowflake divertor

Steady-state At ELM peak


Snowflake divertor is a leading heat flux mitigation candidate for NSTX-U

• Single and double-null radiative divertors and upper-lower snowflake configurations considered– Supported by NSTX-U divertor coils and compatible with coil current limits– ISOLVER modeling shows many possible equilibria

• Impact of changing IOH on snowflake minimal • Reduced divertor coil set can be used for snowflakes

NSTX-U simulation

NSTX-U double-null NSTX-U double-snowflake-plus

NSTX-U double-snowflake-minus


Four divertor coils should enable flexibility in boundary shaping and control in NSTX-U

A variety of lower and both lower and upper divertor snowflake configurations are possible in NSTX-U with four coils per divertor• ISOLVER free-boundary Grad-Shafranov solver used • Four coils can be used to control up to four parameters (X-pts, OSP, etc)

Midplane flux surface

0.0 1.5 mm

3.0 mm

6.0 mm

9.0 mm

Ltot (m) 38.3 20.4 13.9 9.9 8.3

LX (m) 16.5 4.0 2.2 1.6 1.5

Angle (deg.)

1.6 0.8 0.99 3.0 3.9

X 2-3 in plasma wetted surface area and connection length vs standard divertor


Simulations of snowflake divertor configuration inform heat flux mitigation scenarios in NSTX-U

• ISOLVER modeling shows a variety of up-down symmetric snowflake (+/-) configurations possible– Impact of changing IOH on snowflake minimal – Reduced divertor coil set can be used for

snowflakes• UEDGE simulations predict heat flux reduction

with carbon by geometry and radiation• Particle removal by cryopump results in reduced

radiation in snowflake


UEDGE settings

• Single null grid based on LRDFIT equilibria (129x129)• Inertial neutral model is used with recycm=-.5• Fixed fraction Argon is varied from 0.5 to 3.5%• Fixed fraction Neon is varied from 0.5 to 6%• 0.9 < y < 1.055• Dperp = .1 (in core) 0.5 (in SOL) m2/s

– Cubic rise from core boundary to separatrix• Thermal diffusivity ci,e = 1 (in core) 3 (in SOL) m2/s

– Cubic rise from core boundary to separatrix• Recycling = 0.99• Power across y=0.9 surface P90 = 9 MW• Zero flux BC for neutral D at core• Psi=0.9 density value determined by fixing particle flux

through core; based on 60 A for 4 MW NBI power 180 A at P90=9 MW

• 300 A centerstack puff at inner midplane.• No drift effects• Heat flux width analyzed for STD divertor only.

– ~3 mm heat flux width expected; fc=1.0 (5 mm) is not unrealistic


Projections with UEDGE edge multi-fluid model for NSTX-U are optimistic

• Grids extend from psi=0.9 to psi=1.2

• STD grid covers 3.1 cm outside the separatrix at the outer MP

• SNF grid covers 3.4 cm outside the separatrix at the outer MP.

Fluid (Braginskii) model for ions and electrons Fluid for neutrals Classical parallel transport, anomalous radial

transport Core interface:

• PSOL90 = 5; 7; 9 MW D = 0.1-0.5 m2/s ce,i = 1-2 m2/s Rrecy = 0.99 Carbon 5 %


In modeled NSTX-U snowflake configuration magnetic geometry shows clear benefits (cf. standard

divertor)


SOL width studies elucidate on heat flux scaling projections for NSTX-U, ST-FNSF and ITER

• Goal: compare SOL width scalings with models• SOL width scaling

– In NSTX: lqmid ~ Ip

-1.6

• For NSTX-U (2 MA): lqmid = 3±0.5 mm

– Multi-machine database (Eich IAEA FEC 2012):lq

mid (mm) (0.63+/-0.08)x Bpol,MP-1.19

• For NSTX-U (Bp~0.55 T): lqmid ~ 1.3 mm

– Comparison with SOL models• Parallel transport: conductive/convective, cross-field :

collisional / turbulent / drift• Goldston heuristic model: B and curvature drift motion

sets SOL width lSOL ~ (2a/R) ri, Spitzer thermal conduction sets Tsep

– For NSTX-U: lqmid ~ 6 mm

• Diagnostics: IR cameras, MPTS, Langmuir probes– Divertor Thomson scattering (incremental) desirable

BP-2


SOL width and divertor heat flux studies in NSTX inform NSTX-U and ITER divertor projections

• Completed FY 2010 DOE JRT – lq

mid characterized in experiment• independent of PSOL and Bt

• constant with increasing fexp

• contracts with increasing drsep; and Ip: lqmid ~ Ip

-1.6

(DIII-D: lqmid ~ Ip

-1.2 ; Alcator C-Mod: lqmid ~ Ip

-1.0)• lq contracts with increasing lithium deposition

– and modeled• XGC0 modeling reproduced Ip dependence (NSTX: lq

mid ~ Ip-1.0 ; DIII-D: ~ Ip

-0.8 ; C-Mod: ~ Ip-0.3 )

• SOLT modeling: weak lq scaling with Pin; Ip scaling is weaker than observed

• Further studies– Connect lq

mid to pedestal width and edge turbulence

– lqmid data analysis for ITER / future STs

• For NSTX-U parameters, lqmid = 3±0.5 mm 36

PAC27-10 PAC27-11

l (Ip)NSTX 128013


SOL width studies in NSTX elucidate on divertor projections for NSTX-U, ST-FNSF and ITER

• JRT 2010 on heat transport– lq

mid contracts with increasing Ip: lqmid ~ Ip

-1.6

• Comparison with SOL models– Parallel transport: conductive/convective, cross-

field : collisional / turbulent / drift– XGC0 reproduced Ip dependence lq

mid ~ Ip-1.0

– SOLT: Ip scaling is weaker than observed• PSOL, collisionality, LII /R set cross-field transport and

turbulence structure that affect lq

– Goldston drift-based model of SOL flows• Attached H-mode regimes• B and curvature drift motion sets SOL width lSOL ~

(2a/R) ri, Spitzer thermal conduction sets Tsep

– Exploring mechanisms setting steep pressure gradient region and connection to SOL width

• Projections to NSTX-U– lq

mid = 3±0.5 mm– As qpeak ~ Ip and qpeak~PSOL, qpeak ~ 20-30 MW/m2

37

PAC29-14

M. M

akow

ski,

AP

S 2

011

PAC29-12

nstx-u 5 year plan for pedestal, scrape-off layer and divertor physics

Documents